回归饱和设计值得商榷的问题

Discussions on some problems for the regression saturation design

摘要: 从理论和实践上论证了回归饱和设计不能用最小二乘法估计回归模型中的参数；同时也论证了回归饱和设计的试验方案，不能用D优良性判断其优劣，回归饱和D最优设计是不存在的。

Abstract: It has been proved by theory and practice that the parameters of regression model can not be estimated by the least square method in the design of regression saturation and that the experiment plan designed by regression saturation can not be evaluated by the D-optimality. As a result, the D-optimal design of regression saturation is not existed.